Cognitive behavioral therapy (CBT) is an effective treatment for many but not all patients with panic disorder. This raises the question of whether there are types of panic disorder for which CBT is effective, and other types for which it is ineffective. Klein's (1993) suffocation alarm theory suggests two types of panic disordered patient: those with intense dyspnea as a frequent panic symptom (suffocation panickers) and those with little or no dyspnea (nonsuffocation panickers). Klein's theory suggests CBT will be less effective for suffocation panickers compared to nonsuffocation panickers. To test this prediction, 22 unmedicated panic-disordered patients were classified as suffocation panickers (n = 13) or nonsuffocation panickers (n = 9) and received 10 sessions of CBT. Both groups had significant reductions in symptoms from pre- to posttreatment, and gains were maintained at 3-month followup. Groups did not differ in treatment response. At 3-month followup 75% of suffocation panickers, and 50% of nonsuffocation panickers were panic-free, and 75% of suffocation panickers and 63% of nonsuffocation panickers were classified as treatment responders. These results fail to support the prediction from Klein's theory and suggest that panic disorder with intense dyspnea can be successfully treated with CBT.